City Research Online

A Market Model for Controlled Resource Allocation in Distributed Operating Systems

Messer, Alan (1999). A Market Model for Controlled Resource Allocation in Distributed Operating Systems. (Unpublished Doctoral thesis, City, University of London)

[img]
Preview
Text - Accepted Version
Download (11MB) | Preview

Abstract

This thesis explores the potential for providing processes with control over their resource allocation in a general-purpose distributed system. Rather than present processes with blind explicit control or leave the decision to the operating system, a compromise, called process-centric resource allocation is proposed whereby processes have informed control of their resource allocation, while the operating system ensures fair consumption.

The motivations for this approach to resource allocation and its background are reviewed culminating in the description of a set of desired attributes for such a system. A three layered architecture called ERA is then proposed and presented in detail. The lowest layer, provides a unified framework for processes to choose resources, describe their priority and describes the range of available resources. A resource information mechanism, used to support choices of distributed resources then utilises this framework. Finally, experimental demonstrations of process-centric resource allocation are used to illustrate the third layer.

This design and its algorithms together provide a resource allocation system wherein distributed resources are shared fairly amongst competing processes which can choose their resources. The system allows processes to mimic traditional resource allocations and perform novel and beneficial resource optimisations. Experimental results are presented indicating that this can be achieved with low overhead and in a scalable fashion.

Publication Type: Thesis (Doctoral)
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Departments: Doctoral Theses
Doctoral Theses > School of Mathematics, Computer Science and Engineering
School of Mathematics, Computer Science & Engineering
School of Mathematics, Computer Science & Engineering > Computer Science
School of Mathematics, Computer Science & Engineering > Computer Science > Computing
URI: http://openaccess.city.ac.uk/id/eprint/20134

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year

View more statistics